Sensor calibration impacts on dust detection based on MODIS and VIIRS thermal emissive bands

Dust detection using remotely sensed measurements has been one of the challenging problems encountered by atmospheric scientists. MODerate Resolution Imaging Spectroradiometer (MODIS) on the Terra (T) and Aqua (A) platforms have been a versatile sensor for well over 21 and 18 years respectively, and have been extremely useful in the retrieval of aerosol information over the entire globe. The MODIS radiances from the Level 1B in general are expected to be within 5% accuracy in the reflective wavelengths and within 1% in the thermal emissive wavelengths. In this paper, we evaluate the sensitivity of previously developed dust detection technique based on thermal emissive wavelengths, which correspond to MODIS bands 20, 29, 31, and 32 respectively. The Thermal Emissive Dust Index (TEDI) performed very comparably to the traditional Aerosol Optical Thickness (AOT) retrievals by MODIS reflective channels. Since the MODIS Thermal Emissive Bands (TEB) are well calibrated on-orbit using a BlackBody (BB) source, the calibration of these long wave infrared bands is quite robust. As A-MODIS continues to perform well beyond its designed lifetime of 6 years, the instrument has undergone various levels of degradation during its mission time. As a consequence, it is imperative to check the impacts of calibration on the higher-level retrievals. In this paper, we rigorously analyze the sensitivity of TEDI due to the impact of calibration by the afore-mentioned TEB. The perturbation of the dominant (linear) calibration term demonstrated the following: first, there was a correlation in the sensitivity of the TEDI due to the uncertainty in the linear calibration term. Based on a perturbation in the linear calibration term for all aforementioned bands over a range of ±5% yielded the TEDI sensitivity to vary from approximately ?3.2% to about ?3.6%. When considering the uncertainty in each individual band significant changes were observed. The least change was observed for the perturbation in the calibration of band 20 with the TEDI sensitivity and the largest sensitivity in TEDI was observed in the perturbation of band 31 calibration. Thus, in the case of TEDI, noticeable sensitivity due to calibration uncertainty was observed in bands 29, 31, and 32, reiterating the importance of the TEB calibration in these bands. Also, the dust detection scheme based on A-MODIS was successfully transferred to the follow-on sensors such as Suomi (SNPP) and NOAA 20 (N20) VIIRS. The results presented in this paper would be extremely helpful in understanding impacts of calibration on the higher-level products for both current and future missions based on the MODIS heritage. Finally, the work also identifies the importance of radiometric fidelity in maintaining the accuracy of the dust detection. Results presented will show drastic improvement of the Saharan dust detection after the reduction of the electronic crosstalk in the 8.5 µm channel of T-MODIS.     

Experiment for mapping land cover and it’s change in southeastern Sri Lanka utilizing 250 m resolution MODIS imageries

Utilizing freely available MODIS NDVI and Natural color imageries of 250 m spatial resolution produced by NASA, an experiment was made to map land-cover and its change with an emphasis on vegetation cover in southeastern Sri Lanka, which plays a vital role for control of green house gas. For the change detection purpose, 1987 land cover map made by present authors from Landsat MSS image and extensive ground truth survey data was used as the base map. The result of the experiment shows that MODIS data are useful to make a land cover map of 250 m spatial resolution for tropical areas with high cloud coverage like Sri Lanka. It was found that the forest cover decrease amounted as large as 21% in 19 years time span in southeastern Sri Lanka, the prominent forest region of the country. On the other hand homestead/vegetation and mixed vegetation/scrub dominant categories increased by 13.7% and 7.1%, respectively. These changes are considered due to a large clearance of forest areas for agriculture and building houses to accommodate increasing inhabitants.     

Nighttime cloud properties retrieval using MODIS and artificial neural networks

In this work a methodology for inferring water cloud macro and microphysical properties from nighttime MODIS imagery is developed. This method is based on the inversion of a theoretical radiative transfer model that simulates the radiances detected in each of the sensor infrared bands. To accomplish this inversion, an operational technique based on Artificial Neural Networks (ANNs) is proposed, whose main characteristic is the ability to retrieve cloud properties much faster than conventional methods. Furthermore, a detailed study of input data is performed to avoid different sources of errors that appear in several MODIS infrared channels. Finally, results of applying the proposed method are compared with in-situ measurements carried out during the DYCOMS-II field experiment.